[Insight into unplanned readmissions: meaning and possibilities of a classification]

We investigated whether a classification method for readmissions, first developed in the United Kingdom, can be applied to Dutch (re-)admission data and possibly used as an indicator of quality.Retrospective analysis of Dutch administrative hospital admission data supplied to the Dutch National Medical Registration database.Readmissions are urgent inpatient admissions that occur within 30 days of discharge from a previous inpatient admission in the same hospital, with the latter being the index admission. Based on the original UK research, readmissions have been classified into seven categories: (A) complication following index admission; (B) anticipated but unpredictable hospital care; (C) patient or healthcare professional preference; (D) artefact; (E) related to a different body system; (F) related to the same body system; (G) social-economic, psychosocial or other indication. This classification was applied to admission data for 18 Dutch hospitals covering the period 2009-2013.Of a total of 1,633,466 admissions, 120,343 (7.37%) were followed by a readmission. The most common category contributing to readmissions was 'the same body system' (category F, 44.07%), followed by readmissions related to different body systems (category E, 27.85%). The most prominent avoidability categories 'complications following index admission (category A, 17.66%) and anticipated hospital care (category B, 5.31%) ended up in third and fourth place respectively. Finally, the joint share for categories C, D and G was 5.55%. The classification of readmissions differs only slightly between different types of hospitals (university, top clinical, general). Large differences in the classification distribution are found between clinical specialties.The classification system for readmissions can be applied well to Dutch admission data and offers an insight into the patterns of admissions and readmissions. Closer in-depth research may provide more evidence concerning the suitability of this classification system as an indicator of hospital quality.